Analysis of dose-response curves and calculation of agonist dissociation constants using a weighted nonlinear curve fitting program

Abstract

The KA value of a full agonist (Furchgott, 1966) is usually calculated following the transformation of a rectangular hyperbolic relationship between equiactive concentrations into a linear relationship via a double reciprocal plot. Such a transformation will introduce error into the value of KA unless the pre- and postinactivation dose-response curves are also rectangular hyperbolas. Nonlinear curve fitting computer programs, which have been developed to evaluate radioligand binding data, fit data to rectangular hyperbolas and, thus, can be used to fit concentration-response data for pre- and postinactivation curves. Examination of (-)-isoprenaline concentration-response curves for positive inotropic effects in left atrial preparations from the guinea pig, established before and after treatment with the irreversible antagonist Ro 03-7894, indicated a marked deviation from the theoretical rectangular hyperbola (Hill coefficient greater than 1). This was accounted for by the presence of a threshold effect. The present study reports on the use of the computer program LIGAND to fit a rectangular hyperbola to the pre- and postinactivation curves for an agonist, while simultaneously calculating and allowing for the presence of a threshold phenomenon. Since both curves are fitted to rectangular hyperbolas, the KA values can be calculated with greater precision than that possible from double reciprocal plots constructed from uncorrected concentration-response curves.